I have the following model that contains a CFA model with zero inflated measures (manifest) variables. I run the model with separate factors, one for the zero inflated process and one for the counts. This is fitted using the EM algorithm and I get good model fit if no other variables or factors are included into the model.

mplusObject(
  TITLE = "SEM - main model;",
  ANALYSIS = "estimator = mlr;STARTS = 20 5;MITERATIONS=1000;",
  VARIABLE = "COUNT ARE x1 (i) x2 (i) x3 (i);",
  MODEL = "F1 BY x1 x2 x3;
           F2 BY x1#1 x2#1 x3#1;
           F1 WITH F2;
           F1 ON x6 x7 x8;
           F2 ON x6 x7 x8;
           ",
  OUTPUT = "standardized TECH1 TECH8;",
  usevariables = colnames(test),
  rdata = test)

If I add an additional factor G1, I seem to get model fit issues. I have tried different factors (different items fed into G1) but this is a setup problem not a data specific issues:

mplusObject(
  TITLE = "SEM - main model;",
  ANALYSIS = "estimator = mlr;STARTS = 20 5;MITERATIONS=1000;",
  VARIABLE = "COUNT ARE x1 (i) x2 (i) x3 (i);",
  MODEL = "G1 BY x4 x5;
           G1 @1;
           F1 BY x1 x2 x3;
           F2 BY x1#1 x2#1 x3#1;
           F1 WITH F2;
           F1 ON x6 x7 x8;
           F2 ON x6 x7 x8;
           ",
  OUTPUT = "standardized TECH1 TECH8;",
  usevariables = colnames(test),
  rdata = test)

Have I setup this incorrectly?

Model fit error messages:

RANDOM STARTS RESULTS RANKED FROM THE BEST TO THE WORST LOGLIKELIHOOD VALUES

5 perturbed starting value run(s) did not converge in the initial stage
optimizations.

Final stage loglikelihood values at local maxima, seeds, and initial stage start numbers:

           -3062.407  unperturbed      0

4 perturbed starting value run(s) did not converge.



THE MODEL ESTIMATION TERMINATED NORMALLY

     WARNING:  THE RESIDUAL COVARIANCE MATRIX (THETA) IS NOT POSITIVE DEFINITE.
     THIS COULD INDICATE A NEGATIVE VARIANCE/RESIDUAL VARIANCE FOR AN OBSERVED
     VARIABLE, A CORRELATION GREATER OR EQUAL TO ONE BETWEEN TWO OBSERVED
     VARIABLES, OR A LINEAR DEPENDENCY AMONG MORE THAN TWO OBSERVED VARIABLES.
     CHECK THE RESULTS SECTION FOR MORE INFORMATION.

However, looking at the STD results, there is a negative variance and the factor loadings in G1 are way above 1 (x5)?

STANDARDIZED MODEL RESULTS


STDYX Standardization

                                                    Two-Tailed
                    Estimate       S.E.  Est./S.E.    P-Value

 G1     BY
    X4            0.106      0.003     32.911      0.000
    X5             6.097      0.439     13.900      0.000

 F1      BY
    AGG                0.296      0.040      7.403      0.000
    SIB                0.259      0.040      6.459      0.000
    DES                0.292      0.038      7.719      0.000

 F2      BY
    AGG#1              0.719      0.000    999.000    999.000
    SIB#1              0.452      0.096      4.720      0.000
    DES#1              0.777      0.000    999.000    999.000

 F1        ON
    MCHAT              0.352      0.089      3.973      0.000
    HQ_TOT             0.097      0.091      1.066      0.286
    CSHQ               0.270      0.076      3.526      0.000

 F2        ON
    MCHAT             -0.154      0.096     -1.607      0.108
    HQ_TOT             0.114      0.079      1.440      0.150
    CSHQ              -0.300      0.073     -4.123      0.000

 F1      WITH
    F2               -0.999      0.001  -1061.904      0.000

 Intercepts
    x4                 1.792      0.074     24.064      0.000
    x5                 2.297      0.136     16.841      0.000
    x1#1               1.357      0.392      3.467      0.001
    x1                 0.972      0.247      3.943      0.000
    x2#1               1.830      0.492      3.717      0.000
    x2                 0.876      0.244      3.585      0.000
    x3#1               2.245      0.426      5.269      0.000
    x3                 0.855      0.221      3.860      0.000

 Variances
    G1               1.000      0.000    999.000    999.000

 Residual Variances
    x4                0.989      0.001   1443.652      0.000
    x5               -36.175    999.000    999.000    999.000
    F1                0.731      0.081      8.982      0.000
    F2                0.877      0.051     17.184      0.000

Unfortunately, I cannot add data to this post.